Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 13 of 13 for “"Tensor Processing Unit"”.

  1. A tensor processing unit design for FPGA benchmarking

    … network architectures with special purpose units to accelerate the computation. However, these accelerator designs vary according to different design structures and properties. It is essential to understand the efficient FPGA architecture for a specific type of workload. This thesis provides …

    texas Repository record for A tensor processing unit design for FPGA benchmarking (opens in a new tab)

  2. A weight-stationary systolic tensor processing unit : from architecture to physical design

    … domain-specific accelerators such as the Tensor Processing Unit (TPU). This report presents BEVO, a from-scratch TPU implemented in SystemVerilog and carried through the digital Application-Specific Integrated Circuit (ASIC) flow on the TSMC 65 nm node. BEVO exposes a five-instruction …

    texas Repository record for A weight-stationary systolic tensor processing unit : from architecture to physical design (opens in a new tab)

  3. Reclaiming Fault Resilience and Energy Efficiency With Enhanced Performance in Low Power Architectures

    … In this work, a novel AI architecture, Tensor Processing Unit, operating at NTC is thoroughly investigated to tackle the issues hindering system performance. Research problems are demonstrated in a scientific manner and unique opportunities are explored to propose novel design …

    utahstate Repository record for Reclaiming Fault Resilience and Energy Efficiency With Enhanced Performance in Low Power Architectures (opens in a new tab)

  4. SHARED LEARNING AMONG DISTRIBUTED EDGE DEVICES USING CORAL EDGE TPU MACHINE LEARNING ENGINES

    … impossible. With the introduction of the Coral Tensor Processing Unit (TPU), real-time data processing through machine learning models on IoT edge devices is achievable. This research explores splitting a convolutional neural network (CNN) to expose an intermediate layer for fine-tune training. …

    nps Repository record for SHARED LEARNING AMONG DISTRIBUTED EDGE DEVICES USING CORAL EDGE TPU MACHINE LEARNING ENGINES (opens in a new tab)

  5. Approximate Computing and In-Memory Computing: the Best of the Two Worlds!

    … Subsequently, we propose the Approximate Tensor Processing Unit (APTPU), a hardware accelerator that utilizes approximate processing elements to replace direct quantization of inputs and weights in ML models, thereby enhancing performance, power efficiency, and area utilization. The APTPU …

    south-carolina Repository record for Approximate Computing and In-Memory Computing: the Best of the Two Worlds! (opens in a new tab)

  6. Computer-aided design of atomic silicon quantum dots and computational applications

    … frameworks. Among them, the matrix multiply unit (MXU) found in Google’s Tensor Processing Unit (TPU) has been indentified to be architecturally favorable for implementation on the DB platform. This work proposes a DB adaptation of the MXU with logic layouts and clocking configurations …

    ubc Repository record for Computer-aided design of atomic silicon quantum dots and computational applications (opens in a new tab) Download PDF for Computer-aided design of atomic silicon quantum dots and computational applications (opens in a new tab)

  7. Towards Scalable, Private and Practical Deep Learning

    … with the latest accelerators such as Graphics Processing Unit (GPU) and Tensor Processing Unit (TPU), training DL models can take prohibitively long time, therefore training large DL models in a distributed manner is a norm. Massive amount of data is made available thanks to the prevalence of …

    unr Repository record for Towards Scalable, Private and Practical Deep Learning (opens in a new tab)

  8. Development of object detection system to minimise tripping and slipping risk

    Falls within the elderly community are estimated to have cost the Australian Healthcare system approximately $498.2 million in 2001 and are predicted to rise threefold to $1375 million in 2051. A vast majority of these falls are a result of tripping, slipping, and stumbling hazards. These types of …

    vu-aus Download PDF for Development of object detection system to minimise tripping and slipping risk (opens in a new tab)

  9. Photonic Tensor Machine and Multi-level Encoding and Decoding in Wavelength-Multiplexed Photonic Processors

    … Large dense matrix computations, also known as tensor operations, has been the major concern because of its intense demand for fast and efficient computations. Facing the challenge, conventional solutions such as conventional Central Processing Units (CPUs) and field-programmable gate arrays …

    queens Repository record for Photonic Tensor Machine and Multi-level Encoding and Decoding in Wavelength-Multiplexed Photonic Processors (opens in a new tab)

  10. Building Efficient Tensor Accelerators for Sparse and Irregular Workloads

    … hardware accelerator development for computing tensor kernels with high performance and energy efficiency. For example, the Google Tensor Processing Unit (TPU) is a custom hardware accelerator targeting efficient matrix multiplications for Deep Neural Networks (DNNs). However, there are …

    gatech Repository record for Building Efficient Tensor Accelerators for Sparse and Irregular Workloads (opens in a new tab)

  11. Embracing Low-Power Systems with Improvement in Security and Energy-Efficiency

    … environment. And, novel power saving opportunities have been unlocked by characterizing the usage pattern of a baseline TPU with rigorous mathematical analysis.

    utahstate Repository record for Embracing Low-Power Systems with Improvement in Security and Energy-Efficiency (opens in a new tab)

  12. Hardware-aware design, search, and optimization of deep neural networks

    … such as CPUs to special-purpose devices such as Tensor Processing Unit (TPU). High-performance computing systems such as GPUs effectively reduce the computation time of DNNs. Due to the slowing down of Moore's law, the research in developing Domain-Specific Hardware, which excels in its assigned …

    iastate Repository record for Hardware-aware design, search, and optimization of deep neural networks (opens in a new tab)

  13. Knowledge extraction and security in internet of things systems

    … as Single Board Computers (SBCs) or low-cost processing modules are being used in areas such as Smart Agriculture or Vehicles and via wireless (2G, 3G, 4G, 5G, Wi-Fi (Wireless Fidelity), LoRa (Long Range), Zigbee, Sigfox, Bluetooth, Satellite...) and stable connections (LAN – Local Area …

    greece Repository record for Knowledge extraction and security in internet of things systems (opens in a new tab)